Hello 🙂 I had 2 kedro-mlflow questions : 1. Is it possible to log YamlDatasets as artifacts in kedro-mlflow? 2. How do you recommend changing the

run id

dynamically (set its value to a timestamp) ? in the current mlflow.yaml, if the run id is null, a new run is created with a random id. If it has a value (environment variable for example), then kedro mlflow searches for an existing run. We want to create a new run with a unique id of our own: what is the best practice to do so? Thanks a bunch ! ☀️

1. See https://kedro-mlflow.readthedocs.io/en/stable/source/04_experimentation_tracking/03_version_datasets.html#how-to-version-data-in-a-kedro-project Example:

my_dataset_to_version:
    type: kedro_mlflow.io.artifacts.MlflowArtifactDataset
    dataset:
        type: yaml.YAMLDataset  # or any valid kedro Dataset
        filepath: /path/to/a/LOCAL/destination/file.yaml # must be a local file, wherever you want to log the data in the end

2. The unique identifiers are managed by MLflow itself, you can have influence on when the generation happens, but I'm pretty sure you cannot influence the run id format because it's created by the MLflow Server / API.

Hi, thanks a bunch @marrrcin for the very quick response ! 2. Works, I thought it was possible to set custom run_ids but I guess not 😊 1. I have exactly this structure but my yamls don't appear in the artifacts (and i get no errors)

Are they appearing in the specified path locally?

yes, they are there 🙂

Yes marrcin syntax is correct, but keep in mind it only works for output artifact. For those which are inputs of a node, there is no out of the box solution, but it is something often required

ah ok ! so for it to be logged it needs to be the output of a node! that explains it : all our yamls are inputs